SEX DIFFERENCES IN THE TABP AND OCCUPATION

Coronary Heart Disease(CHD) accounted for 60% of deaths by Cardiovascular Disease(CD) in Australia (Heart Facts R e p o r t , 19 88 ) . Sex differences emerge in that CHD is a more frequent cause of death for women than for men. Women also have a greater incidence of hypertension, an established CHD risk factor. In so far as the TABP confers coronary risk then the issue of sex differences in TABP incidence and

manifestation a r i s e s . Price (1982) provides a

cognitive/social learning model of the TABP which accounts for both individual and population differences in TABP manifestation. Price argues that it is important to study such differences to obtain a better understanding of the TABP itself. The simplicity of the TABP label, chosen by

Friedman and Rosenman as a neutral term to minimise cross- disciplinary controversy, may imply a degree of homogeneity which cannot be justified.

Although the Framingham Study provided early evidence on the manifestation of the TABP amongst women and the issue of male/female differences the TABP has nonetheless been less systematically studied in the female population (Baker, Dearborn, Hastings & Hamberger,19 84; Gree n g l a s s , 1987 ; Kelly & Houston, 1985; Lawler & Schmeid, 1987; Waldron, 1977 ) and less often (Weidner, 1989 ) than in the male population. The Framingham Study (Haynes, Feinleib & Kannel,1980) found that

women aged 4 5 to 64 with CHD had higher TABP

s c o r e s ,suppressed hostility, tension and anxiety scores than women without CHD . Type A women developed twice the rate of CHD and three times the rate of angina as Type B women. Type A women either in paid outside employment or as housewives were at a greater risk for CHD than their Type B counterparts. Type A housewives had CHD rates similar to those of Type A employed women. Cross-sectional studies provide additional evidence that the TABP is related to the incidence of CHD in women,e.g., Keningsberg, Zyzanski, Jenkins, Wardwell & Licciardello (1974). Waldron (1983) concluded that the TABP can increase women's risk status for cardiovascular disease(CD) with coronary atherosclerosis of particular concern.

Further impetus for the study of the TABP in women and of male/female differences stems from the convergence in the

rate of CHD for both sexes in older age and the fact that middle-aged women are much more likely to die from infarction than males (Thoresen & Low,1990). There is in addition more general concern for women's health in a changing society in which women are exposed to increasing stress, especially as a response to multiple roles ( Cooper & Davidson,1981; Froberg, Gjerdingen & Preston,1986). While women may benefit from role diversification role strain remains an issue (Froberg et al, 1986).

The relative neglect of women as subjects in TABP research reflected the concern of the 1950's and '60's for the predominance of men in cardiac wards. The consequence of this "male" emphasis is well expressed by Thoresen & Low

( 1990 ) :

" Simply put, we do not at present have a very clear understanding, scientifically as well as clinically, of TA in women" (p 118) .

In an earlier review Baker et al (1984) noted the methodological and design problems evident in many studies. These included inadequate reporting of statistics, the use of very small samples, unsystematic subject selection and self-selected samples. Large-scale epidemiological studies over-interpreted small differences owing to their large numbers of degrees of freedom. Measurement issues were ignored with, for example, different TABP scales treated as if they were interchangeable, overlooking the possibility of the scales addressing different aspects of the TABP notion.

Thoresen & Low (1990), reviewing the literature some years later, are no more positive or complimentary. To quote:

" Unfortunately, the research on TA in women has been plagued by inadequate assessment. With rare exception, measures used with women have not been validated against

significant outcomes over time in female populations" (p 119) .

Measurement issues appear to be of primary importance. The Structured Interview(SI), considered the most valid predictor of CHD (Booth- Kewley & Friedman,1987), may not be especially applicable to women. The Si's failure to generate a continuous scale (Thoresen & Low,1990) and its (stereotyped) male oriented item content (Matthews & Haynes,1986) are two criticisms. The SI measures overt signs of hostility, angry behaviour and speech stylistics found less often in women than in men. Its emphasis on competitive behaviour and its work orientation may also render it less useful as an assessment tool with female subjects (Matthews & Haynes,1986).

Sex differences in reactivity appear to exist at the physiological level. Frankenhauser, Dunne & Lundberg (1976) exposed males and females to two stressors, a cognitive task and repeated venipuncture. They found that both groups responded with increased heart rate and feelings of distress. Between the sexes, however, there were different patterns of adrenaline excretion with the females not

responding with elevated adrenaline as did the males. Other mechanisms appeared to underlie the females' increased responsiveness.

High Type A people often show greater autonomic arousal than Type B's in some situations (Williams,1985). Men display more obvious cardiovascular responses to challenge and competition than women who react with lower increases in heart rate and blood pressure (Lawler & Schmeid,1986). The type of challenge used experimentally, however, is important. Tasks such as mental arithmetic fail to produce measurable changes (Anderson,Williams, Lane, Haney, Simpson & Houseworth, 1986) yet MacDougall, Dembroski & Krantz (1981) found that Type A women reacted more strongly than Type B women to verbally challenging interactions ( a history quiz ) . Thus the salience of the stressor to women is a factor influencing research outcomes. Finally Polefrone & Manuck (1988) report two further sources of possible sex differences in reactivity. The magnitude of differences between Type A and Type B women seems to be less than for men. Secondly hormonal changes may influence reactivity levels. Hastrup & Light (1984), in an earlier study, found that normally menstruating women in the middle to later follicular phase of their cycles showed lower heart rate and systolic responses to a stressful task than men. Women in the middle of the luteal phase were no different

Thoresen & Low ( 1990 ), in their review of sex differences, chose a number of personality variables in an analysis that will be partially summarised here owing to its pertinence to the present study. These authors firstly examined possible links between hostility and CHD in women but were unable to draw any definite conclusions. Hostility is defined as an attitudinal set consisting of suspiciousness, cynical distrust and resentful thoughts. Anger is the emotional expression of hostility. Thoresen & Low comment that the constructs " Anger-In" and Hostility from the Framingham Study appear confused although the study did find a positive relationship between Anger-In and CHD for women.

The issue of hostility, anger and sex differences in TABP requires further clarification. How anger is expressed is an issue for high Type A's who, possessed of hostile attitudes (the cognitive component), become emotionally reactive (angry) and behave aggressively ( insultingly or abusively ), thus damaging relationships. Shope (1978) found that women use verbal aggression where men increase both physically and verbally aggressive responses to perceived insult. McCann, Woolfoik, Lehrer & Schwarcz (1987) reported that women turn anger inwards more than men and that a significant relationship existed for women between the TABP and suspicion.

At this stage it should be recalled that some authors argue that aggression is a central component of the TABP,

e.g., Weekes & Waterhouse ( 1990 ). Yet it is an area where evidence exists for a genuine sex difference, biologically determined(Henry & Stephens,1977). To the extent that a measuring instrument taps into the aggressive component of the TABP it may yield a misleading result that men are more Type A than women.

Complicating hormonally determined sex differences in aggression is the issue of sex role norms, suggested by McCann et al (1987) as an important topic for future research. Price(1982) in her cognitive/social learning model sees women as equally subject to the internal beliefs that underlie Type A behaviour. Aggressive behaviour may therefore be more appropriately viewed as "masculine sex- role" determined rather than as a central, core TABP component. Auten, Hull & Hull (1985), in a study supportive of this argument, found that higher Type A scores were associated with a masculine sex-role orientation, as measured by the Bern Sex- Role Inventory, regardless of gender. The implication for a conceptualisation of the TABP is that aggressive behaviour is a "distractor" , diverting attention from more important underlying ( cognitive ) mechanisms which determine beliefs and attitudes. The study of sex differences may have contributed to a clearer ( more "pure" ) definition of the Type A Behaviour Pattern.

Earlier sections reviewed possible links between CD, stress and the TABP. Following Byrne & Byrne's (1990)

analysis it was noted that stress was too vague a term. Consequently the more widely accepted notions of anxiety, depression and neuroticism were discussed. Thoresen & Low

(1990) use a similar approach in reviewing sex differences. They suggest the possibility of a stronger link between the TABP and anxiety, depression and neuroticism for women than for men. It is also possible that " fight or flight" reactions could contribute to CHD in highly aroused Type A people in that such a response is easily triggered in this population. A picture begins to emerge of the TABP as "disorder of arousal" with pathogenic implications for cardiovascular health if, as Thoresen & Low (1990) speculate, " the mechanism for atherosclerosis includes increased plasma catecholamine levels, increased blood pressure reactivity and greater changes in cardiac function( e.g., heart rate,stroke volume )" (p 125).

High Type A women report more stress, frustration and anxiety, poorer health, lower self-esteem, more nervousness and dysporia (Dearborn & Hastings,1987; Haynes & Feinleib,1980; Kelly & Houston,1985; Waldron,1978). Some cross-sectional studies, e.g., Bass(1984), Booth-Kewley & Friedman (1987), report links between CHD and these symptoms yet do not permit firm conclusions owing to the possible confounding effects of other variables as well as other methodological weaknesses (Thoresen & Low,1990). Once again sex-roles enter the analysis with DeGregorio & Carver (1980) finding an association between the above symptoms and a

"feminine" sex-role orientation for high Type A women. This is scarcely a surprising result given the more general finding that women as a group report depression and anxiety more often than men (Henderson, Byrne & Duncan-Jones,1981) . The suppression of anger amongst Type A women is no less a health issue than amongst women generally.

Developmental research using child and adolescent samples serves to shed some light on the sex differences issue. In addition it should be noted that the disease process of atherosclerosis begins in childhood, developing slowly over time (Weidner,1986). Bergman & Magnusson (1986) conducted a fourteen year longitudinal study and concluded that the TABP could be identified at a young age and was stable over time. Aggression was more applicable to the male sample and motor hyperactivity to the female sample. There were no sex differences in the adult Type A scores, leading the authors to conclude that different components of the TABP were reinforced for males and females early in life, ultimately leading to a similar outcome. Bergman & Magnusson concede, however, that a measure of the TABP was applied only at the conclusion of the study owing to the unavailability of a suitable measure at the study's outset, a fact which must bring the study's validity into some question.

Matthews (1984) reviewed the issue of the TABP and children. She concluded that the setting of escalating standards and parental disapproval play an important

aetiological role, at least for males.Matthews argues that the literature on achievement motivation emphasises sex differences more clearly, revealing different parent/child interaction patterns leading to the development of the TABP according to the child's sex. Aggressive behaviour and hostility are not equally encouraged between males and females resulting in aggression becoming a fairly stable characteristic in males but not in females (Matthews,1984 ).

In reviewing the issue of sex differences in the TABP Baker et al ( 1984 ) conclude that " the majority of studies suggest that adult females are no more or less Type A than males, especially when other identified correlates are controlled. As in male samples, clear- cut, positive correlations were repeatedly found between Type A behaviour and occupation, education and socio-economic status " (p

483). Thoresen & Low (1990) are more tentative, commenting that " a definitive coronary-prone behaviour pattern for women is yet to be identified and confirmed via controlled, prospectively designed studies " (p 127).

The importance of studying the TABP in the occupational setting has already been noted. Burke & Weir (1980) found that in male administrators there was a positive relationship between TABP scores and hours worked, levels of concentration, responsibility accepted, level of overload, rate of organisational change, stress in communicating and total stress from work conditions. Type A's also reported that their jobs negatively affected their home lives.

Research with women yields similar results (Kelly & Houston,1985; Waldron,1978) although a more recent study by Bedeian, Mossholder & Touliatos (1990) found a relationship between the TABP and working longer hours held only for men. Hartel & Chambless (1989) found that for both full-time and part-time working women prevalence rates for the TABP were the same as for men. Byrne & Reinhart (1990) reported no sex differences in a sample of health workers except on the Job

Involvement sub-scale of their measure, the JAS. Controlling for occupational status, however, reduced the sex difference substantially owing to the over- representation of nem at higher levels. The Type A Pattern may have similar psychological and behavioural correlates in women as it does in men( Greenglass, 1990 ) although further research is needed before firm conclusions can be drawn

(Bedeian et al, 1990).

Waldron (1977) found that women who enter the workforce score more highly on the TABP than those at home. Price (1982) states that women at home may be subject to pressures that promote Type A behaviour in that domestic duties lack clear, unambiguous performance criteria. Women who are also in the workforce, however, are subject to further stressors which may enhance a Type A predisposition. While it may be, as Waldron ( 1980 ) points out, that high Type A women are more likely to enter the workforce as an expression of their Type A status, pressures associated with the workplace are likely to exacerbate the pattern (Greenglass,1990). Two

likely sources of pressure are the reduced amount of free time available and the demands imposed through performing multiple roles (Greenglass,1990).

The issue of role conflict-induced stress and sex differences needs to be viewed within a broader context of the changing pattern of female participation in the workforce. More women work outside the home than ever before (LaCroix & Haynes,1987). Women are entering the job market at levels previously occupied by men. Patterns of female participation in the workforce are changing with women taking on a greater variety of roles (Lawler,1987). More women are aspiring, for example, to supervisory positions, raising issues such as coping styles and the possibility of interactions between sex and the TABP in control decisions (Greenglass, 1987 ; Miller, Lack & Asroff, 1985 ). In consequence concern has been expressed not only for women's health in a period of rapid social change (Baker et al, 1984; Froberg et al, 1986 ) but also for such issues as discrimination and inequalities of access to promotion and higher status positions (LaCroix & Haynes,1987; Gutek,1988).

Issues of "equal employment opportunity" are especially relevant to a consideration of sex differences in the TABP in the workplace. Haynes & Feinleib (1980) found CHD rates to be double for women in clerical jobs than for housewives. Women, relative to men, are over-represented in lower status jobs characterised by tediousness and monotony (Price,1982). They are often powerless, lacking in autonomy and control

and usually receiving little recognition for their efforts (Lawler & Schmeid,1987). Even when women hold similar positions to men they are subject to added occupational stressors such as pay inequities and sex discrimination

(Greenglass, 1987 ).

An examination of the part played by socio-economic factors yields some interesting conclusions. Shekelle, Schoenberger & Stamler (1976) found no sex differences in the TABP when socio-economic status was controlled. A more recent study by Moss, Dielman, Campanelli, Leech, Harian, Van Harrison & Horvath (1986) using the Structured Interview produced a similar result. Greenglass (1990) concluded that such factors as occupational and educational status should be included in any study of TABP sex differences as well as noting reported differences between women at varying occupational levels. Waldron (1977) found that women of higher educational status and working full-time versus part- time had higher TABP scores.

In summary, a number of conclusions may be drawn which serve to guide the direction of the present study:

1) The occupational environment, an important setting for the expression and manifestation of the TABP, may operate to reduce or eliminate any TABP sex differences that may have existed prior to employment.

2) The absence of sex differences is more readily apparent when major socio-economic variables are controlled.

The two most prominent variables are educational and occupational status.

3) Given the argument that control issues are very salient to high Type A individuals then the over­ representation of women in low-control/high demand jobs becomes an important issue.

4) Women employed outside the home may be rendered more vulnerable to the development of psychological symptoms than men through the mechanism of a pre-existing Type A predisposition being aggravated in a setting that does not provide an outlet for the TABP' s expression. It is also conceivable that a formerly Type B individual may develop more Type A reactions under the same circumstances.

5) To the extent that Type A reactions result in prolonged psychological distress symptoms which may in turn convey greater cardiovascular disease risk then concern for women's health in this context is appropriate. The suggestion that women may express themselves differently to men, for example the expression of anger, appears especially relevant in its implications for the development of symptomatology.